Talker-specific adaptation: how listeners learn and use indexical information during speech processing

speech perception;talker adaptation;eye-tracking;episodic memory;amnesia;sleep consolidation;accents

Trude, Alison

关键词: speech perception;    talker adaptation;    eye-tracking;    episodic memory;    amnesia;    sleep consolidation;    accents;   

Others  :  https://www.ideals.illinois.edu/bitstream/handle/2142/46681/Alison_Trude.pdf?sequence=1&isAllowed=y

美国|英语

来源: The Illinois Digital Environment for Access to Learning and Scholarship

【 摘 要 】

Humans’ ability to understand speech is remarkable in that, despite large amounts of inter-talker variability due to factors such as pitch, speech rate, and accents, we are usually able to understand what is being said quickly and with little conscious effort. However, there is still much to be understood about the processes by which we learn about talker-specific information in the speech signal and the memory mechanisms that support this learning. In this dissertation, I present a series of seven experiments examining listeners’ on-line processing of a novel foreign accent and the contributions of the declarative memory system and sleep-dependent consolidation in learning talker-specific information. In the first series of experiments, participants’ eye movements were monitored as they listened to the speech of a native Québec French speaker who spoke with an accent that should have made it easier to disambiguate the names of the images in the display. Despite highly accurate performance at identifying the target words, listeners’ eye movements revealed difficulty when listening to the French talker. However, analyses examining learning across the course of each experiment showed that participants did improve as they gained more exposure to the accent. I conclude that talker adaptation does not always happen rapidly, and that experience with a particular accent is crucial. In the second set of experiments, I explored the memory mechanisms responsible for talker adaptation by testing amnesic patients with severe declarative memory impairments and by using manipulations of sleep in healthy undergraduate participants. Both studies used an eye-tracking paradigm in which participants heard a regional accent of American English. Amnesic participants performed much like healthy comparisons, using accent information to facilitate understanding. This finding suggests that episodic memory is not necessary for talker-specific learning. I also examined healthy, college-aged participants’ performance on this eye tracking task in two sessions, with a period of intervening sleep or wakefulness. Participants who slept performed better overall at Session 2, suggesting that sleep-dependent consolidation processes can aid in the process of spoken word recognition. Taken together, the results of these experiments extend our knowledge of the time course and memory mechanisms that support the learning of talker-specific information.